Você está na página 1de 6

IJRET: International Journal of Research in Engineering and Technology ISSN: 2319-1163

__________________________________________________________________________________________
Volume: 02 Issue: 03 | Mar-2013, Available @ http://www.ijret.org 276
LOW POWER SRAM USING ATD CIRCUIT

Himanshu J. Shah
1
, Veerendrasingh Tiwari
2

1, 2
SAGAR INSTITUTE OF RESEARCH & TECHNOLOGY, Bhopal, MP
hims.shah67@gmail.com, virendrasingh1180@gmail.com

Abstract
The address signal lines are connected to a semiconductor memory device when they are implemented on the PCB (Printed Circuit
Board). So they experience the wiring capacitance and due to this wiring capacitance on printed circuit board the rising and falling
time of respective address signal disperse. If the address signal lines change in timing, the following problem occurs. When a
microprocessor connected to a semiconductor memory device changes the address information from one address to another address,
due to deviation of timing of the change of the signal, the wrong address is generated. In an Asynchronous semiconductor memory
device, the change of address is immediately responded. So it also responds to such wrong address information and internal circuit
selects the wrong address information and operates corresponds to that wrong address.

Key words:- OFDM, STBC, SWTICHING, PATH FADING.
-----------------------------------------------------------------------***-----------------------------------------------------------------------
1. INTRODUCTION
Based on the operation type of individual data storage cells,
RAMs are classified into two main categories. Dynamic RAMs
(DRAM) and Static RAMs (SRAM). The DRAM cell consists
of a capacitor to store binary information, 1 or 0, and a
transistor to access the capacitor. Cell information is degraded
mostly due to a junction leakage current at the storage node.
Therefore, the cell data must be read and rewritten periodically
(refresh operation) even when memory arrays are not accessed.
On the other hand, the SRAM cell consists of a latch, therefore,
the cell data is kept as long as the power is turned on and
refresh operation is not required. The classification of
memories is as shown in following chart.



Figure 1.1: Classification of Memory

Read-only Memory (ROM) allows, as the name implies, only
retrieval of previously stored data and does not permit
modifications of the stored information contents during normal
operation. ROMs are nonvolatile memories. Depending on the
type of data storage (data write) method, ROMs are categorized
as Mask ROM, in which data are written during chip
fabrication by using a photo mask, and Programmable ROM
(PROM), in which data are written electrically after the chip Is
fabricated. Depending on data erasing characteristics, PROMs
are further classified into Fuse ROM, Erasable PROM
(EPROM) and Electrically Erasable PROM (EEPROM). The
data written by blowing the fuse electrically cannot be erased
and modified in Fuse ROM. DATA in EPROMs and
EEPROMs can be rewritten, but the number of subsequent re-
write operations is limited to. Flash memory is similar to
EEPROM, where data in the block can be erased by using a
high electrical voltage. A drawback of EEPROM is the slower
write speed, in the order of microseconds. Read write (RW)
memory must permit the modification (writing) of data bits
stored in the memory array, as well as retrieval (reading) on
demand. The read write memory is commonly called Random
Access Memory (RAM), mostly due to historical reasons.
Unlike sequential-access memories such as magnetic tapes, any
cell can be accessed with nearly equal access time. The stored
data is volatile; i.e., the stored data is lost when the power
supply voltage is turned off. A system based on this approach
consists of one or more subsystems which are surrounded by an
environment with which they communicate. The information
and state of the system is usually held in storage elements and
is allowed to flow between those during the duration of a
discrete time interval. In synchronous design the boundaries of
this time interval are physically represented by the rising and
falling edges of a clock signal generated by a global clock
generator. These edges are used to trigger the storage elements
to store the new information on their inputs. Since the
information must be stable before storing it, the length of the
time interval is determined by the worst case performance of
the slowest operation of the system. Unfortunately this forces
operation that is faster and thus completes early to idly wait for
IJRET: International Journal of Research in Engineering and Technology ISSN: 2319-1163

__________________________________________________________________________________________
Volume: 02 Issue: 03 | Mar-2013, Available @ http://www.ijret.org 277
the next occurring clock edge. The storage elements are often
represented by registers in the form of flip-flops or transparent
latches. Designing a test memory core an its peripherals: The
task is to be executed by designing an asynchronous SRAM
core (4 * 4) row and column decoders and current amplifier.
Designing an ATD circuit for above said memory core. Testing
the ATD circuit for functional verification with the SRAM
memory core.

2. ASYNCHRONOUS RAM AND ADDRESS TRANSITION
DETECTOR
Such an asynchronous semiconductor memory device
immediately follows change of the address signals A0-Am and
performs the decoding operation at the row decoder RD and the
column decoder CD. A single word line (not shown) in the
memory cell array is selected by the row selection signal X and
a single data on plurality of bit lines (not shown) is selected by
the column selection signal Y, and thereby such data is output
through the I/O gate and data output buffer DB0 . The row
decoder RD of a number of Fig. 2.3 is composed of the decoder
circuit



Figure 2.1: Schematic Diagram of Conventional Decoder
Circuit

such as shown in Fig. 2.1, the number being equal to the
number of row selection signals X0 - Xk, and each decoder
circuit outputs selection signal when the respective combination
of the specified address signals is input.



Figure 2.2: Timing Diagram For Conventional Asynchronous
Semiconductor Memory

In Fig. 2.1, Q20 is a depletion type MOS transistor, Q0-Qi are
enhancement type MOS transistors. When all of the input
address signals become low level, the row selection signal X
becomes high level. In addition, the column decoder CD is also
configured in the same way unwanted power consumption as
mentioned above is generated particularly in the input stages of
the row and, column decoders as is explained with reference to
Fig. 2.2. For convenience of explanation, it is supposed that the
3-bit address signals A0-A2 are input. Fig. 2.2 indicate the
address signals A0-A2 and address data a-d composed of
combination of A0-A2 and outputs Xa - Xd of the decoder
circuit corresponding to address data a-d. In a process
wherein an external device connected to a memory device
changes address data from address a to address b, for
example, the address signal A1 changes, and in its transitional
condition, the address c is generated in unwanted form and it
is then decoded. When an address changes in such unwanted
form, a certain decoder circuit operates as shown in Fig. 2.2e.
And a certain decoder circuit operates as shown in Fig 2.2g.
But another decoder circuit operates in unwanted way as shown
in a Fig.2.2f and it is unnecessary to provide such output.
Accordingly, unwanted AC power consumption occurs in the
memory device, increasing total power consumption. Such
power consumption is also generated in the case wherein a
noise signal as indicated by P in Fig. 2.2a is added to the
address signal line, even when an external device does not
change address data. In this case, the unwanted address d as
shown in Fig. 2.2d is generated; an output Xb of the decoder
circuit which is not required to change may change as shown in
Fig. 2.2g, and an output Xd of the decoder circuit which has
decoded the address d also changes as shown in Fig. 2.2h.
Thus unwanted power consumption occurs. When an output of
the decoder circuit changes, the circuit connected to the
IJRET: International Journal of Research in Engineering and Technology ISSN: 2319-1163

__________________________________________________________________________________________
Volume: 02 Issue: 03 | Mar-2013, Available @ http://www.ijret.org 278
succeeding stage thereof operated following such change and
unwanted power consumption also occurs as explained above.
This is the major problem which occurs when dealing with the
asynchronous semiconductor memory

3. ATD CIRCUIT
In the case wherein a semiconductor memory device is
mounted on a printed circuit board, the rising time and falling
time of respective signals on a plurality of address signal lines
connected to a semiconductor memory device disperse due to
the dispersion of wiring capacitance on a printed circuit board.
If signals on a plurality of address signal lines change with
some scattered deviation in the timing, the following problem
occurs.



Figure 3.1: ATD CIRCUIT

To avoid above said problem, the change in address signal is
detected and hold the internal circuit to a non-operative
condition in response to any change of address signal for a
specified period. And this is accomplished by the extra circuitry
which is called ATD (Address Transition Detection) circuit
which generates the pulse of a specified period when one of the
address signal changes. This circuit relates to an asynchronous
semiconductor memory device and more specifically to a
semiconductor memory device which has reduced power
consumption as a result of holding an internal circuit to a non-
operative condition in response to any change of address signal
for a specified period following the change of said address
signal

4. CIRCUIT LEVEL DESIGN AND FUNCTIONAL
VERIFICATION
Figure 4.1 and 4.2 shows the basic block diagram for the
interfacing of memory peripherals and memory core with ATD
circuit and without ATD circuit which is used for functional
verification of the ATD circuit.


Figure 4.1: Interfacing Of Memory Peripherals and Memory
Core without ATD Circuit

Each and every block shown in the Figure 4.1 and 4.2 is
implemented in the cadence tool using UMC 180 library used
for 180 nm technology and the value of the supply voltage
VDD for 180 nm technology is 1.8 volt.



Figure 4.2: Interfacing Of Memory Peripherals and Memory
Core with ATD Circuit

4.1 IMPLEMENTATION OF ATD CIRCUIT
For the design of ATD circuit is first of all it is required to
design the delay circuit. And here in place of delay circuit a
single buffer or the chain of buffer is used. And for that inverter
design is require which explain as below.
IJRET: International Journal of Research in Engineering and Technology ISSN: 2319-1163

__________________________________________________________________________________________
Volume: 02 Issue: 03 | Mar-2013, Available @ http://www.ijret.org 279


Figure 4.1.1: Schematic of The Circuit For Inverter

To achieve switching threshold Vm = VDD/2 = 0.9 volt and
also to reduce the glitch at the transition of the input, the size of
the pMOS (W/L)p = 12.375 and for nMOS it is (W/L) n =
1.5.



Figure 4.1.2: Transient And DC Analysis For Inverter

Here, from the transient response of the inverter shown in
Figure 3.12, it is observed that the switching threshold Vm is
achieved to VDD/2 that is 0.9 volt by keeping the size of
transistors as mention earlier. The critical voltage VIL, VIH,
VOL and VOH is as given below, VIL = 0.795 volt, VIH =
1.032 volt VOL = 0 volt and VOH = 1.8 volt. From these
values the noise margins NML and NMH can be calculated.

NML = VIL - VOL = 0.795 0 = 0.795 volt

And NMH = VOH - VIH = 1.8 1.032 = 0.768 volt

The rising delay (tr) and falling delay (tf ) and also tpLH and
tpHL of inverter for the unloaded and loaded condition is
discuss along with buffer.


Figure 4.1.3: Schematic Of The Circuit For XOR Gate

Figure 4.1.3 shows the schematic of XOR gate using CMOS in
which the size of all the transistors used is decided from the
minimum sized inverter that designed earlier. The size of each
pMOS (W=L)p used in XOR gate is 24.75 and for nMOS it is
3. In the XOR schematic of Figure 3.20, the value of Wn is 180
nm and length of each transistor is also Wn. So W/L ratio of
each transistor is same as the multiplication factor with the Wn
which is indicated besides each transistor.



Figure 4.1.4: Transient And DC Analysis For XOR Gate

Figure 4.1.4 shows the transient response and the DC analysis
of the XOR gate. The transient response shows the two
different inputs of the XOR gate with 3.33 MHz and 1.66 MHz
of frequencies and the output of the XOR gate which is high
only when both the inputs are unequal (01 and 10) and the DC
analysis shows the switching threshold of the XOR gate which
is 0.9 volt. tpLH, tpHL and p for the output of XOR circuit for
different loaded condition is mention in table 4.1below.


IJRET: International Journal of Research in Engineering and Technology ISSN: 2319-1163

__________________________________________________________________________________________
Volume: 02 Issue: 03 | Mar-2013, Available @ http://www.ijret.org 280
Table 4.1: Transient Analysis For XOR Gate



4.2 SIMULATION RESULTS
Figure 4.1 shows the SRAM array which has 4 *4 (16) SRAM
cells with row and column decoder which are without ATD
circuit. According to the status of the address lines and R!/W,
reading or writing operation can be performed. Writing
operation is done through BL. According to address line status
the corresponding output of decoder goes high. And word line
status depends on the R!/W which select the complete row of
the cell and for particular selection of cell the output of decoder
goes high depends on inputs address of column decoder. So the
reading or writing operation can be performed to that particular
cell. Written or readied data can be seen through Q.



Figure 4.1: Schematic of the Circuit For 16 Cells With Row
And Column Decoder


Figure 4.2: Transient Response for SRAM With ATD Circuit

Figure 4.2 shows the transient response of the reading and
writing operation to the cell with introducing the ATD circuit to
the decoder. In which address line A1 has low level for all the
time and a pulse generated due some noise or parasitic
capacitance introduce in the circuit. it is recognized as the
wrong address 01 which should not be detected by the
decoder. And due to that wrong address detection the data store
in the cell 2 is changed which is shown by output of cell 2 (Q2).
Now, Figure 4.2 shows the response with introducing the ATD
circuit. Now in this circuit it is observed that the D2 RD is not
go high for the row decoder address input combination of 01
and the status of cell 2 not changed. So the wrong address
which generated in the previous circuit is eliminated by
introducing the ATD circuit in the decoder which avoids the
wrong data to be writing to the second cell of SRAM core.
Here, by introduction of ATD circuit, any address which of the
period less than 152 ns is considered as the noise and it not
detected. So, for this design the address change frequency
IJRET: International Journal of Research in Engineering and Technology ISSN: 2319-1163

__________________________________________________________________________________________
Volume: 02 Issue: 03 | Mar-2013, Available @ http://www.ijret.org 281
should be less than 3.26 MHz. so, this ATD circuit is designed
for the processors 8085 (8 bit) which operates on the frequency
of 3.07 MHz, 8086 (16 bit) which operates on the frequency of
5 MHz and the PIC controller (PIC12F629/675 - 8 bit) which
operates on the frequency of DC - 20 MHz.

CONCLUSIONS
In this work, an ATD circuit was designed to suppress the
effect of noise on the address line in cadence virtuoso
environment. The ATD circuit was sound to support a address
change frequency up to 3.26 MHz. The testing was done on a
4*4 asynchronous SRAM core and the functionality of the
circuit was successfully verified. The test circuit of 4*4 SRAM
core was also designed in same environment. All the
simulations were done for 180 nm technology node.

REFERENCES
[1] Providing e-Transaction Guarantees in Asynchronous
Systems with No Assumptions on the Accuracy of Failure
Detection Paolo Romano and Francesco Quaglia,
JANUARY-FEBRUARY 2011
[2] Dr. Jackson Hu, Recent Business Model and technology
Trends and Their Impact on the Semiconductor Industry,
IEEE Asian Solid State conferences , Nov 2007.
[3] 256 Mb x32 Synchronous DRAM, Micron Technology
Inc., 2003
[4] Current Sensing Completion Detection In Dual-Rail
Asynchronous Systems. Lukas Nagy & Viera
Stopjakova,2012
[5] Jan M. Rabaey & Anantha Chandrakasan, Digital
Integrated Circuits A Design Perspective,2nd ed., Pearson
Education, Inc., as Pearson Prentice Hall, 2009.
[6] Shiv Shankar Mishra, Adarsh Kumar Agrawal and R.K.
Nagaria, A comparative performance analysis of various
CMOS design techniques for XOR and XNOR circuits,
International Journal on emerging Technologies, Fab 2010.
[7] Hisatada Miyatake, Ohtsu (JP),Asynchronous
Semiconductor Memory, United States Patent, Prior
Publication Data, Sept. 1, 2009.
[8] Formal Specification and Runtime Detection of Dynamic
Properties in AsynchronousPervasive Computing
Environments Yiling Yang, Yu Huang, Member, IEEE,
Jiannong Cao, Senior Member, IEEE, Xiaoxing Ma,
Member, IEEE, and Jian Lu,2012

Você também pode gostar